Interactions, organization, and emergent behaviors are fundamental to all complex systems. These characteristics can be observed in every real-world scenario, prompting the question: how do these interactions among elements serve as the foundation of complex systems? Over the years, researchers have developed various methods to answer such a question. One widely used bottom-up approach is agent-based simulation, which deduces a system’s properties by examining its components. Alternatively, a system can be modeled via a top-down approach, such as hypergraphs, mathematical structures abstracting higher-order interactions. These two methods provide different perspectives on the same system, each offering unique and valuable insights. In this Ph.D. project, my primary focus is on social systems, aiming to shed light on the dynamics of group behaviors in online social media.